Orbit, our internal CRM augmented with AI
A complete CRM we designed, built and use every day: scheduling, marketing automation, support, project tracking, quoting and invoicing, sales deck generation and an in-meeting AI assistant.
ClientMPWB (projet interne)
An internal product, proven on ourselves
Orbit is not a technical demo: it is the tool our business runs on. The booking module you meet elsewhere on this site is part of it.
Building it in-house holds us to the same standard we apply for a client — reliability, usability, scale — with immediate feedback whenever a product decision turns out badly.
Pulling a business back together from ten tools
An agency rarely runs on one piece of software: one tool for scheduling, another for invoicing, a third for support, a fourth for project tracking. Every boundary between two tools is a place where information gets lost or retyped.
The goal was not to rebuild those tools one by one, but to put them on a shared data model: a client, a project and an invoice have to mean the same objects everywhere.
One platform, with AI where it actually helps
Orbit covers the full cycle, from booking a call to tracking an invoice. AI is not cosmetic here: it works on specific tasks where the gain is measurable, mainly meeting write-ups and project analysis.
Scheduling and marketing automation
Booking, reminders and follow-up sequences run off the same database as the rest of the client relationship, with no syncing between third-party tools.
Quotes, invoices and project tracking
Quoting, invoicing and project progress share one source of record. Moving from accepted quote to open project to invoice requires no re-entry.
Sales deck generation
From meeting write-ups and notes, Orbit produces a structured sales document. The formatting time after a call becomes marginal.
AI project analysis
Client projects are analysed to surface weak signals — scope drift, blockers, untouched topics — before they turn into problems.
Client portal
Clients follow their project progress and invoicing on their own, from the same data foundation the internal team uses.
AI assistant for Google Meet
The assistant captures meeting content and writes it back into the CRM, attached to the right client and project, with no manual rework.
The technologies we chose
Laravel for the application core, a Vue.js interface, PostgreSQL and Redis for data and job queues, language model APIs for the AI features, and the Google Meet API for the meeting assistant.
Laravel
Vue.js
PostgreSQL
Redis
LLM APIs
Google Meet API
A tool that doubles as our proving ground
Orbit lets us validate in real conditions the approaches we later recommend: targeted AI integration, client portals, invoicing automation. What we advise, we run ourselves first.
Working on something similar?
Give us thirty minutes. We will tell you plainly what is feasible, at what budget and on what timeline.